Talk:CrossCrypt

Compatibility with loop-AES and Twofish loopback encryption
It seems to me that the original author has been misled into believing that CrossCrypt containers are always implicitly compatible with loop-AES, and thus suffer from severe flaws. However, their web page merely states that "SingleKey Mode aes-loop compatible" which sounds like while it does provide legacy support for loop-AES, it has to be explicitly enabled. I'll dubious these statements for now. -- intgr 12:18, 12 December 2006 (UTC)
 * Never mind, the web site documents "multikey support" on its TODO list as full loop-AES compatibility. -- intgr 15:05, 19 December 2006 (UTC)

Biased comparison
The CrossCrypt section is heavily biased towards TrueCrypt, I tagged it with pov. Note that "CBC" may also mean CBC with ESSIV, which does not share the vulnerabilities of plain CBC. I can't find any sources about which mode CrossCrypt really uses. -- intgr 12:24, 12 December 2006 (UTC)
 * I second that. I'd wote for removing the whole section. If those two storage encryptions where the only ones, or just the most used, there'd be a point in having a section devoted to a comparision of them. But with the vast number of different implementations, I'd rather have an article comparing all of them. How to do such an article is a bit tricky. We could compare the different pieces of software, but we could also compare the different formats, or different modes. (Formats and modes are kind of orthogonal, one format can have a header field allowing different modes, and two different formats could use the same mode). A comparision of implementations could mention which formats they support, but even that would not be sufficient, as it might not implement all features supported by the format.
 * Since the article says CrossCrypt is compatible with loop-AES, I think it is true that predictable IVs are used. Calling this mode CBC is not completely correct, as CBC means the IV is chosen randomly. The claims about plausible deniability of TrueCrypt might be exaggerating, but I believe it is a bit better than CrossCrypt in this respect. Kasperd 06:48, 19 December 2006 (UTC)
 * I'll remove the section.
 * I've, too, been thinking of writing a "comparison of" article on the topic. Will probably do it sooner or later, if someone else doesn't beat me to it. -- intgr 14:58, 19 December 2006 (UTC)